Detection and classification of high frequency transients using wavelet analysis

This paper assesses the use of wavelet analysis to detect and classify high frequency transients. Work within the field of wavelet analysis is reviewed and the applications within the field of power system condition monitoring are explored. High frequency transients arising from lightning and switching phenomena are simulated using ATP/EMTP. The feasibility of using wavelet analysis to detect and classify high frequency transients is investigated. Results show that though detection is possible with the use of very fast fault recorders, classification using data from wavelet decomposition at a single level is difficult. This paper indicates that the distribution of energy across a number of levels leads to the accurate classification of high frequency transients. The use of wavelet analysis together with an effective classification procedure can be the basis for the development of a fast and accurate online monitoring technique.